Вы находитесь на странице: 1из 17

Organization Science

Vol. 23, No. 1, JanuaryFebruary 2012, pp. 177193


ISSN 1047-7039 (print) ISSN 1526-5455 (online)

http://dx.doi.org/10.1287/orsc.1110.0650
2012 INFORMS

Regions Matter: How Localized Social Capital Affects


Innovation and External Knowledge Acquisition
Keld Laursen
DRUID, Department of Innovation and Organizational Economics, Copenhagen Business School,
2000 Frederiksberg, Denmark, kl.ino@cbs.dk

Francesca Masciarelli
Department of Management and Business Administration, University G. dAnnunzio,
65127 Pescara, Italy, f.masciarelli@unich.it

Andrea Prencipe
Department of Management and Business Administration, University G. dAnnunzio, 65127 Pescara, Italy;
and SPRUScience Technology Policy Research, University of Sussex, Brighton BN1 9RH, United Kingdom, a.prencipe@unich.it

o introduce new products, firms often use knowledge from other organizations. Drawing on social capital theory and
the relational view of the firm, we argue that geographically localized social capital affects a firms ability to innovate
through various external channels. Combining data on social capital at the regional level, with a large-scale data set of the
innovative activities of a representative sample of 2,413 Italian manufacturing firms from 21 regions, and controlling for a
large set of firm and regional characteristics, we find that being located in a region characterized by a high level of social
capital leads to a higher propensity to innovate. We find also that being located in an area characterized by a high degree
of localized social capital is complementary to firms investments in internal research and development (R&D) and that
such a location positively moderates the effectiveness of externally acquired R&D on the propensity to innovate.
Key words: social capital; social interaction; external R&D acquisition; internal R&D; product innovation
History: Published online in Articles in Advance May 17, 2011; revised August 10, 2011.

Introduction

innovation, whereas in a geographically dispersed network, membership per se does not affect innovation. In
addition, Alccer and Chung (2007) analyze the conditions under which firms may choose to locate in regions
with high levels of spillovers while taking account of
the increased probability of knowledge leakage to competitors in the regions. Although these contributions are
very valuable, they do not model the social mechanisms
that transmit knowledge in geographical locations, and
they do not examine the effects of these mechanisms
on firm-level outcomes, such as innovation, in general
(however, see Owen-Smith and Powell 2004).
We help fill this gap in the literature by introducing
the notion of geographically localized social capital to
explain outcomes in the form of product innovation. We
posit that geographically bound social capital is the key
transmitter of knowledge spillovers within geographically constrained areas and that the resulting existence
of localized social capital has implications for firms
abilities to innovate. Specifically, this paper is, to our
knowledge, the first to identify geographically localized
social capital as a key factor in promoting firm-level
innovation and to provide quantitative evidence to support the claims made. The extant empirical literature
on knowledge sourcing suggests that firms rely heavily on knowledge exchange with external parties, such

The ability of firms to introduce product innovations


is acknowledged to be a key determinant of organizational performance. As long ago as 1890, Marshall
argued that geographical proximity promotes knowledge
spillovers that benefit firms knowledge production
i.e., positive externalities in the form of ideas that are
taken up by others and combined with suggestions of
their own; and thus becomes the source of yet more new
ideas (Marshall 1890, p. 332). In other words, the production of knowledge in the local geographical environment affects the ability of firms to introduce innovations
based on new ideas. Following Marshall (1890) and
Jacobs (1969), various studies have analyzed knowledge
spillovers and highlighted their localized nature (e.g.,
Alccer and Chung 2007, Almeida and Kogut 1999, Bell
and Zaheer 2007, Gambardella and Giarratana 2010,
Jaffe et al. 1993, Owen-Smith and Powell 2004, Sorenson 2003). In a seminal quantitative empirical study,
Jaffe et al. (1993) find that citations in patents tend to
come from the same geographic areas and that the intensity of knowledge spillovers is uneven across regions
because of different region-specific mechanisms. OwenSmith and Powell (2004) show that membership in a
geographically colocated network (Boston+), based
on formal contractual relationships, positively affects
177

178

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

as suppliers, customers, universities, other key individuals, and sometimes even competitors (Chesbrough 2003,
Landry et al. 2002, Laursen and Salter 2006, von Hippel
2005). The central argument in this paper is that geographically bound social capital facilitates joint learning
for innovation and reduces the search and transaction
costs of both contractual and noncontractual interaction among the economic actors in a region. We define
regional social capital as the localized norms and networks that enable people to act collectively within a
region. This conceptualization follows Woolcock and
Narayans (2000, p. 226) general definition of social capital. We examine a key dimension of social capital
namely, the level of social interaction at the regional
level.
This paper draws on social capital theory, the relational view of the firm, and the geography literature to
examine the importance of regional social capital for
firms innovative capabilities. Using these elements, we
build a theory that predicts a positive effect of regional
social capital on firm-level innovation outcomes, and
we explain how investments in both internal and externally acquired research and development (R&D) interact with regional social capital in explaining firm-level
innovation. Our empirical investigation exploits withincountry regional variation to investigate the effects of
social capital on innovation. We use the case of Italy
because we have access to regional data and also to
reliable microeconomic data on firms and their innovative activities. Italy also provides an interesting test case
because (a) social capital is unevenly distributed across
Italian regions (Putnam et al. 1993), (b) Italy constitutes an empirical focus for prominent sociology inputs
to the social capital debate (Banfield 1958, Putnam et al.
1993), and (c) Italy features prominently in the tradition of research on industrial districts (e.g., Brusco 1982,
Piore and Sabel 1984). To analyze the effect of social
capital on product innovation, we combine data on social
capital at the regional level with a large-scale data set on
innovative activities in a representative sample of 2,413
Italian manufacturing firms from 21 regions. We show
that being located in a region characterized by a high
level of localized social interaction leads to a higher
propensity to innovate. This result holds after controlling for a large set of firm and regional characteristics,
as well as regional political participation.1 We find that
firms that have invested in absorptive capacity in the
form of internal R&D are more likely to benefit from
localized social interaction. Moreoverand in line with
our proposed theory of localized social capitalwe find
that social capital in terms of localized social interaction
represents an external contingency that positively affects
the efficiency of the external sourcing of R&D; i.e., location in an area characterized by a high degree of social
interaction positively moderates the effect of externally
acquired R&D on innovation outcome.

Organization Science 23(1), pp. 177193, 2012 INFORMS

In terms of the originality of our contribution, our


study is novel in introducing social capital into the literature on geographically localized knowledge creation:
to the best of our knowledge, this is the first study
that demonstrates the existence of Marshallian knowledge spillovers linked to localized social capital. One
set of important contributions in the management literature (e.g., Nahapiet and Ghoshal 1998, Tsai and
Ghoshal 1998) focuses on social capital and innovation in intrafirm relationships, but as the authors in
this tradition themselves acknowledge, their analyses
can be extended to other institutional settings characterized by enduring relationships, including interorganizational interactions. Another pertinent contribution
(Owen-Smith and Powell 2004) addresses the issue of
firms network positions in local networks and firmlevel innovation and emphasizes the strategic benefits
of formal connections, but does not address the role
of ties to individuals or the social aspects of collaboration. We also extend the literature on geographically
localized social capital (e.g., Kalnins and Chung 2006,
Uzzi 1997) to encompass firms product innovation more
explicitly by explaining how firms investments in internal and external R&D are intertwined with localized
social capitalderived from the social interaction of
individualsin producing innovation outcomes, and we
provide a set of micromechanisms that underpin the
macroconcept of localized social capital in the context
of innovation. As a starting point, we posit that interaction on innovation between a focal firm and its environment has two essential requirements: the exchange
of information/knowledge, and the provision of trust to
support joint activities in a highly imperfect market. We
add to the social capital literature by specifying two theoretical effects of localized social capital that facilitate
these requirements: geographically localized connectivity and trust effects. On a related note, we extend the
argument made in the innovation literature that external
knowledge acquisition can be an important ingredient
in product innovation beyond the focus on the relationship between internal and external knowledge sources
and its effect on firms abilities to introduce product
innovation (Arora and Gambardella 1990, Cassiman and
Veugelers 2006, Schmidt 2010). Given pervasive market failures in the markets for technology (Arora et al.
2001)and holding the interaction between internal and
external R&D constantwe introduce localized social
capital as an important contingency for effective external
R&D investments.
Finally, we make an important empirical contribution
in considering social capital as a factor that is embedded in the local geographical environment of the firm
not something linked to its narrow external relationships,
e.g., associations with suppliers, customers, and universities. In following the original idea by Putnam et al.
(1993) to link social capital to the local geography, we

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation


Organization Science 23(1), pp. 177193, 2012 INFORMS

avoid a setup where observed social capital could be


nothing more than a reflection of a focal firms ability to establish well-functioning external relations. Our
approach, on the other hand, creates a desirable distance
between the independent and dependent variables.
Empirical and Theoretical Background
The importance of social variables to explain differences
in economic outcomes across regions has a long tradition. Banfield (1958) argued that southern Italys economic backwardness was due to a lack of social capital,
while Putnam et al. (1993) has found that the performance of Italian social and political institutions is influenced strongly by citizens engagement in community
affairs, i.e., by social capital. The study by Putnam and
his colleagues prompted a substantial body of research
on the relationship between social capital and economic
outcomes (Nahapiet and Ghoshal 1998) as well as many
empirical studies focusing on the role of geographically
constrained social capital at different levels of aggregation, ranging from cities (e.g., Jacobs 1961) and regions
(e.g., Beugelsdijk and van Schaik 2005, Putnam et al.
1993) to entire nations (e.g., Knack and Keefer 1997,
Zak and Knack 2001).
The relational view of the firm prompts the conjecture that firms critical resources often span firm
boundaries and are embedded in interfirm resources
and routines (Dyer and Singh 1998); therefore, complementary resource endowments, effective governance,
and external partners are important sources of knowledge for new ideas and information that can result
in product innovations and consequential supernormal
profits. The Marshallian view posits that agglomeration is driven by spillovers of information, knowledge,
and new ideas. Spillovers are context-dependent as a
result of two interrelated factors. First, an important
component of knowledge is tacit, difficult to unbundle from its context (sticky), and complex; therefore, its transfer requires information-rich, face-to-face
interactions (Nelson and Winter 1982, Szulanski 1996,
von Hippel 1994). Second, personal contactsthe major
means of face-to-face interactionare fostered by proximity and are less likely to be established over larger
geographical distances (Rosenthal and Strange 2003,
Storper and Venables 2004). In the words of Rosenthal
and Strange (2003, p. 387), Information spillovers that
require frequent contact between workers may dissipate
over a short distance as walking to a meeting place
becomes difficult or as random encounters become rare.
Given that individuals are proximate within geographical regions, these geographic spaces play a key role in
defining the geographic boundaries of social capital. In
addition, through their social interactions, the individuals
residing within particular regions develop shared identities based on industry similarities and associated communication codes, which facilitate cooperation within
these individual regions (Romanelli and Khessina 2005).

179

Our analysis rests on three additional assumptions.


First, personal relationships overlap work relationships
because social capital is dependent on individual attitudes and behaviors, which impinge on the collective
behavior of firms. This assumption is supported by the
literature on localized economic activities, which argues
that multiple-level networks (professional and personal)
eventually merge within geographical locations (Brusco
1982, Saxenian 1994). Saxenian (1994) argues that the
success of Silicon Valley compared with other regions,
such as Route 128, is based on a more robust exchange
of ideas among firms and other local institutions, facilitated by a system of collaboration and learning. Second,
we take location as a given. To a degree, this assumption is supported by Dahl and Sorenson (2009), who
find that entrepreneurs place much more emphasis on
closeness to family and friends when deciding where
to locate those businesses than on regional characteristics that might influence the performance of their ventures. Third, we assume that social capital is effective
as a factor that facilitates the transmission of information and creates trust, irrespective of the regional level
of development (we control for the level of development
in the empirical analysis). As North (1989) suggests,
once economic relations extend beyond the local level,
transaction costs related to monitoring and enforcement
increase markedly, and the local social network has to be
replaced and/or complemented by formal organizations
and institutions. There is evidence of a substitution effect
between social capital on the one hand and other institutions associated with higher levels of development and
overall more efficient markets on the other hand (see,
for instance, Guiso et al. 2004). Although this evidence
implies that social capital is more important in underdeveloped areas, this is not definitive. For instance, Knack
and Keefer (1997) demonstrate that trust and civic norms
are stronger in countries with formal institutions that
enable effective contract and property rights protection,
and La Porta et al. (1997) suggest that public and private
institutions are less effective in countries with low levels
of trust among citizens.

Hypotheses
Social Interaction and Innovation
The innovation literature demonstrates that firms innovation processes depend strongly on external actors, e.g.,
users, suppliers, universities, and competitors (see, e.g.,
Arora et al. 2001, Shan et al. 1994, von Hippel 1988).
The distributed or open (Chesbrough 2003) nature of the
innovation process is derived from its information and
knowledge requirements: innovation necessitates combinations of a variety of new and existing knowledge bases
located inside and outside the focal firm. We develop the
hypothesis that high levels of geographically bounded
social capital, in terms of social interaction in the home

180

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

region, may generate competitive advantage for local


firms in the form of innovation, because localized social
capital favors information and knowledge flows among
firms and external actors within regions.
The idea central to social capital theory is that high
levels of social interaction provide information benefits
in terms of access, i.e., the opportunity to obtain a valuable piece of information; of timing, i.e., the opportunity
to be informed early; and of referrals, i.e., having your
name prominent at the right time and in the right place
(Burt 1992). The central contributions to social capital
theory (as expressed in the work of, e.g., Ronald Burt
or Sumantra Ghoshal) have been cast mostly in rather
general terms and do not relate to geographic proximity. However, some authors within the sociological
network tradition emphasize the important role of geographical proximity in the context of innovative activities (e.g., Owen-Smith and Powell 2004, Stuart and
Sorenson 2003). Stuart and Sorenson (2003, p. 232)
describe this proximity in the context of innovation as
follows: When people with common professional interests cluster in physical space, informal social and professional networks emerge and serve to disseminate information. Indeed, as Uzzi (1997, p. 62) suggests, some
of the benefits of geographically localized networks is
that they allow for the transmission of face-to-face interaction channeling fine-grained information that allows
later information processing and problem recognition.
Given that regional social capital is defined in terms
of norms and networks, it favors innovation because it
helps to connect people across different organizations
and to combine their knowledge components within particular regions. We term this the localized connectivity effect. Whereas explicit knowledge may be relatively
easy to obtain through minor efforts, such as reading
journals or benchmarking, social interactions enable a
closeness between firms that facilitates the exchange of
the deeper, tacit components of knowledge (Kogut and
Zander 1996, Lane and Lubatkin 1998). An implication
of the relational view (Dyer and Singh 1998, Lane and
Lubatkin 1998) is that social interaction among knowledge workers within a region also may enhance the
ability of firms to recognize and evaluate local external
knowledge, providing better access to and understanding of specialized information, language, know-how, and
the operations of other actors and allowing more efficient communication. Localized social interaction not
only can improve flows of knowledge from the supply
side, but it also can function as a channel to enhance
firms understanding of the demand side, enabling a better knowledge about (local) user needsa factor that
has been found to be critical for innovation success
(Rothwell et al. 1974, Slater and Narver 1994).
The social interaction component of social capital supports access to informal channels of knowledge within
a region. Following Granovetter (1973), Burt (1992)

Organization Science 23(1), pp. 177193, 2012 INFORMS

argues that the strength of weak ties stems from the


possibility of these ties bridging otherwise disconnected
groups of individuals and firms. Such relations can lead
to boundary-spanning searches, often seen as necessary for successful innovation (Fleming and Sorenson
2001, Rosenkopf and Nerkar 2001). Provided that localized (in our case, intraregional) networks are effective
for facilitating innovation, localized social capital in
terms of social interaction may ease the process of an
external knowledge search by providing a richer set of
communication channels (Sorenson 2003, Sorenson and
Audia 2000), thereby increasing the chances that problem solvers with complementary knowledge can make
fruitful connections with respect to innovation.
Not only does localized social capital connect knowledge workers through connectivity effects, but it also
improves the functioning of knowledge connections by
alleviating potential moral hazard problems through the
creation of trust. We term this the localized trust effect.
Transaction cost theory suggests that a solution to problems related to information asymmetries would be the
use of formal safeguardsfor instance, in terms of
hostage exchange of assetsand third-party interventions for conflict resolution (Pisano 1990, Williamson
1979). In contrast, the relational view of the firm suggests informal safeguards as alternatives or complements
to formal safeguards (Dyer and Singh 1998). Informal
safeguards rely on trust and embeddedness as well as
related reputation effects.
Previous research suggests that informal safeguards
are more effective for complex exchanges, such as those
related to innovation (Dyer and Singh 1998, Hill 1995,
Uzzi 1997). Central to this argument is that repeated
social interaction leads to increased trust (Granovetter
1985, Gulati 1995, Tsai and Ghoshal 1998). This implies
that localized, regional social interaction would lead
to localized trust, which is in line with Saxenians
(1994) findings for Silicon Valley. According to Gulati
(1995), there may be at least two reasons why repeated
social interaction increases trust. First, as firms interact via the interactions of their employees, they learn
about each other and develop trust based on shared
notions of fairness (Gulati terms this knowledge-based
trust). Second, trust arises from repeated interactions:
the behavior of one firm may be perceived as untrustworthy by another local firms workforce, leading to costly
sanctions that exceed the potential benefits of opportunistic behavior (Gulati terms this deterrence-based
trust). Reports of noncooperation are likely to spread
rapidly within a geographically bounded area, such as a
region, as a result of the high degree of social connectedness in that location. Regardless of how trust is created
within a geographically proximate area, it is likely to
increase both formal and informal knowledge exchange
among firms, and to increase the likelihood that the firms
will be innovators. Because product innovation is the

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation


Organization Science 23(1), pp. 177193, 2012 INFORMS

result of a combination of internal and external knowledge, we suggest that innovation is affected by social
connections and consequential trust within regions; i.e.,
Hypothesis 1. Firms operating in regions with high
levels of social capital in terms of social interaction are
more likely to introduce product innovation.
Social Capital and Internal R&D
Previous research suggests that there is complementarity
between internal and external knowledge sources (Arora
and Gambardella 1990, Cassiman and Veugelers 2006),
i.e., that the return from one variable increases with
increases in another variable. In our case, we expect
that (1) firms own investment in R&D will increase
the value of social capital for producing innovative output, and (2) the value of R&D investments will increase
with higher levels of regional social capital. The first
expectation is grounded in the literature on absorptive
capacity (Cohen and Levinthal 1990): to benefit from
external knowledge transmitted through social capital
and the localized connectivity effect described previously, firms must invest in in-house knowledge. Without such investment, it will be very difficultif not
impossibleto identify, assimilate, and exploit knowledge from the external environment. In other words, not
all firms can be expected to benefit equally from regionally bound social capital in the form of social interaction.
Those that invest more in R&D should benefit more from
location in an environment endowed with high levels of
social capital.
The second expectation is based on the idea that
in-house R&D may not provide sufficient inspiration or variety to enable combinations of knowledge
required to produce innovation (Rosenkopf and Almeida
2003, Rosenkopf and Nerkar 2001). Thus, a combination of in-house and beyond-firm boundary search is
required (Rothaermel and Alexandre 2009). Rosenkopf
and Nerkar (2001, p. 292) state that the gains associated with the internal development of technology are
not sustainable unless the organization is able to integrate external developments. Note that our focus in this
paper is on organizational boundary spanning through
the localized connectivity effect exerted by the focal firm
on other agents within the region, but it does not preclude the importance of boundary spanning outside the
region. However, out-of-region organizational boundary
spanning is more difficult to handle in the case of innovation, given that it is likely to involve less face-to-face
interaction because this is more costly over longer geographical distances (Morgan 2004).2
In sumand factoring in both absorptive capacity and
localized boundary-spanning argumentswe hypothesize the following.
Hypothesis 2. Social capital and internal R&D
spending are complementary in affecting the likelihood
of introducing product innovation.

181

Social Capital and Externally Acquired R&D


There are several reasons why investment in externally
acquired R&D might fail to facilitate innovation in the
acquiring firm.3 In this context, research on external
knowledge acquisition is informed by transaction cost
economics (Pisano 1990; Teece 1986, 1988; Williamson
1985). Because R&D projects are complex and involve
asset-specific investments, R&D contracts are hard to
define ex ante (Arora et al. 2001), introducing imminent
risks of hold up (Oxley 1997, Williamson 1979). The
relational view of the firm posits that social capital may
work as an institutional reparation mechanism for the
implied market failures, facilitating solutions to ex ante
and ex post transaction cost problems (Dyer and Singh
1998). Following this more general logic, high levels
of regional social capital should enhance intensive and
repeated interactions, as well as the creation of trust,
reciprocity, and mutual expectations among the actors
in the region. In addition, social interactions develop
over time in dyadic relationships as formal exchange
partners become more comfortable with and confident
about each others competencies and reliability in economic exchange (Larson 1992, Ring and van de Ven
1994). Repeated interactionsincluding localized geographical interactionmay augment the actors incentives to exchange information relevant to transforming
outsourced R&D into innovations (Dyer and Singh
1998, Larson 1992, Zahra et al. 2000), and recurring
social interactions facilitate the sharing of expectations
and goals and reduce the need for formal monitoring
(Yli-Renko et al. 2001). Certainly, regions that are characterized by extensive social interactions may favor the
sharing of expectations and reduce the need for formal
monitoring within the region. Indeed, these localized
factors increase the probability that outsourced R&D
will result in product innovation.
Reduced transaction costs is not the only determinant of external knowledge acquisition, however.
The resource-based approach highlights the process of
resource accumulation and learning in the decision
to acquire external knowledge (Robins and Wiersema
1995). Specifically, successful outsourcing may require
outsourcing experience and involve trial and error. In
fact, the effectiveness of external R&D acquisition
depends on the ability of the firm to recognize and assess
the value of external knowledge and to understand the
willingness of other actors to share useful information
(Dyer and Singh 1998, Yli-Renko et al. 2001). Obtaining this ability through learning on the part of acquiring
firms may be facilitated by geographical locations characterized by high levels of social interaction and derived
trust (Brusco 1982, Lundvall 1992).
In sum, high levels of regional social capital generate an environment that facilitates the process of search
for complementary knowledge and increases trust among
the parties involved through localized connectivity and

182

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

trust effects. Operating in a region with high levels of


social capital may provide better opportunities to learn
how to deal with the management of outsourced R&D
activities. In short, if two otherwise equal firms invest
the same amounts in outsourced R&D, the firm located
in a social capital-rich region is more likely to produce an innovation than the firm in a social capital-poor
region. Accordingly,
Hypothesis 3. The effectiveness of externally acquired R&D on the likelihood of introducing product
innovation is higher for firms operating in regions associated with high levels of social capital.

Data Description

This research uses firm- and region-level variables,


constructed from data from different databases. The
firm-level data on innovation come from Capitalia (an
Italian banking group), which collects data on Italian
manufacturing firms through stratified random sampling
of manufacturing firms with more than 10 employees
(Capitalia 2006). The survey refers to the three-year
period 20012003, and the sampling plan was created
by subdividing the population of assigned firms into layers (strata). The population from which the sample was
extracted consists of approximately 70,000 firms, representing about 7% of the total number of firms and
9% of the total employees. The survey was based on
a questionnaire instrument administered through telephone interviews and achieved a response rate of 28.5%.
The final sample is representative of Italian manufacturing firms across four macroregions (i.e., northwest,
northeast, center, and south), Pavitts (1984) sectors (i.e.,
supplier-dominated, scale-intensive, science-based, specialized supplier), and firm sizes (1120, 2150, 51250,
251500, more than 500 employees) (Capitalia 2006).
The number of observations with no missing values for
any of our variables is 2,413 firms.
The regional-level data used to analyze structural
social capital were collected by the Italian National Institute of Statistics (ISTAT) through multiscope analyses in
1999. Response to this survey, which was based on telephone interviews, was compulsory: the response rate was
82.5%. Individual responses were aggregated by ISTAT
at the level of the 21 Italian regions. To validate the
geographical unit of analysis (the 21 regions) against
the alternative of the Italian provincial level (regions
were selected to include at least 2 provinces, and the 21
regions chosen represent 103 provinces), we conducted
a components of variance analysis with random effects
(see, e.g., McGahan and Porter 1997) for three variables
related to geographic structural social capital, which
are available also at the provincial level: (1) per-capita
gross domestic product (GDP) for 2001, (2) participation rate in the 2001 referendum on institutional reform,
and (3) per-capita legal protests for lack of payment

Organization Science 23(1), pp. 177193, 2012 INFORMS

of obligations for 2001. For these three variables, we


find that the regional level accounts for between 68%
and 77% of the total variance, whereas the provinciallevel accounts for between 18% and 26% (the residual represents 5%6%). Accordingly, we believe that
the level of the 21 Italian regions is the most relevant
level of aggregation for our purposes, because variation in social capital levels is more likely to be between
rather than within regions. To measure regional expenditure on R&D as a percentage of regional GDP, regional
human capital, and population size, we use data from
EUROSTAT, the European Commission (EC) statistical
office, for 1999. The units are the regions corresponding to the Nomenclature of Territorial Units for Statistics
level 2 (NUTS 2), the classification adopted by the EC.
To an extent, we avoid the problem of common
method bias because our dependent variables are collected at firm level, whereas some of the key independent variables (in particular, those related to social capital) are collected at the individual level and aggregated
at the NUTS 2 regional level. To reduce the effects
of consistency artifacts for the firm-specific variables,
the questions related to outcome variables were positioned in the survey after the questions related to the
independent variables (Salancik and Pfeffer 1977). We
performed a Harmans one-factor test on firm-level variables on the models in this paper to examine whether
common method bias might be augmenting the relationships detected (Podsakoff and Organ 1986). Because we
found multiple factors, and because the first factor did
not account for the majority of the variance (the first
factor accounts for only 18% of the variance), the test
does not indicate common method bias.
The Structural Social Capital Measure
The issue of how to measure social capital and identify its sources and consequences is contested (Portes
and Landolt 1996). Nahapiet and Ghoshal (1998) and
Lindenberg (1996) propose distinguishing between the
structural and relational dimensions of social capital. The structural dimension refers to informal social
interactions amongst individuals; the relational dimension refers to the assets rooted in those relationships
(e.g., trust and trustworthiness). Following mainstream
research on social capital (e.g., Coleman 1988, Portes
1998, Putnam et al. 1993, Woolcock and Narayan 2000),
we consider the structural dimension to be the most
suitable for empirical analysis because it focuses on
the sources as distinct from the consequences of social
capital. The structural dimension is also considered the
most suitable for an empirical analysis because it is
much easier to achieve an accurate measure. To measure the structural social capital of the Italian regions,
we selected a total of 10 regional social capital variables (see Table 1). We include variables for friendships
and spare-time socialization (Meeting friends regularly,

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

183

Organization Science 23(1), pp. 177193, 2012 INFORMS

Table 1

Description of the Variables Included in the PCA

Variable

Description

Participation in cultural associations

People age 14 and older who have joined meetings in cultural circles and similar ones
at least once a year in the 12 months before the interview, for every 100 people of
the same area

Participation in voluntary associations

People age 14 and older who have joined meetings in voluntary associations and
similar ones at least once a year in the 12 months before the interview, for every
100 people of the same area

Participation in nonvoluntary organizations

People age 14 and older who have joined meetings in nonvoluntary organizations
at least once a year in the 12 months before the interview, for every 100 people of
the same area

Voluntary associations per region

Number of voluntary organizations for every 10,000 people

Meeting friends regularly

People age 6 and older meeting friends at least once a week, for every 100 people of
the same area

Social meetings

People age 6 and older attending bars, pubs, and circles at least once a week in the
12 months before the interview, for every 100 people of the same area

Satisfaction as to relationships with friends

People age 14 and older who are satisfied with their relationships with friends

Unpaid work for political parties

People age 14 and older who have carried out unpaid work for a political party in the
12 months before the interview, for every 100 people of the same area

Money given to parties

People age 14 and older who have given money to a political party at least once a
year, for every 100 people of the same area

Participation in political meetings

People age 14 and older who have joined a political meeting in the 12 months before
the interview, for every 100 people of the same area

Social meetings, and Satisfaction as to relationships with


friends), participation in social organizations (Participation in cultural associations, Participation in voluntary
associations, Participation in nonvoluntary organizations, and Number of voluntary associations per region)
and participation in political movements (Unpaid work
for political parties, Contributions to political parties,
and Participation in political meetings).
We ran a nonparametric principal component analysis (PCA) on the social capital variables. This differs
from the standard PCA and derives eigenvalues from
a cograduation matrix (Spearmans rho or rank-order
correlation coefficients). The aim is to minimize the
effect of outliers. Table 2 presents the two principal
components we extracted from the analysis. These two
components explain 81.6% of the total variance. This
is considered a very satisfactory result for the analysis of social variables. The first factor captures social
interaction, and the second captures political participation. The social interaction component, which is of key
interest, appears to capture important aspects of individuals social networks consistent with our adopted definition of social capital and the social capital literature; the
variable for political participation is a control variable
because of its ambiguous effect in the Italian context.
Specifically, and in line with Putnams proposed measurement of social capital for social interaction, we consider items that capture regional participation in informal
associations (Putnam et al. 1993) and socialization with
friends (Putnam 2000). Both items reflect the breadth of
social ties essential in social capital theory.

Table 2

Matrix of Factor Loadings

Variable
Participation in
cultural associations
Participation in
voluntary associations
Participation in
nonvoluntary organizations
Number of voluntary
associations per region
Meeting friends regularly
Social meetings
Satisfaction as to
relationships with friends
Unpaid work for political parties
Money given to parties
Participation in political meetings

Component 1:
Social
interaction

Component 2:
Political
participation

0877

0324

0890

0222

0928

0267

0745

0380

0827
0908
0861

0104
0055
0039

0410
0184
0631

0838
0880
0693

The individual items underlying the social capital


social interaction construct can be linked to geography in
the following way. For the three items for participation
in organizations, our reasoning is based on the fact that
all these items relate to participation in physical meetings (see Table 1). These meetings are almost certain
to have been within-region meetings. For the item on
the number of associations per region, this necessarily
relates to a region. Based on insights from the sociological literature (e.g., Sorenson and Audia 2000), we would
argue that friendships also are most likely to be local,
i.e., within-region.

184

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

Econometric Analysis
Measures
Dependent Variables. The dependent variable for the
first logit analysis is a dummy variable that takes the
value of 1 if the firm introduced a product innovation and 0 if it did not. This variable is based on the
responses to the following question: Over the threeyear period 20012003, did your firm introduce product
innovation, by which we mean did the firm introduce
at least one new (or a significantly improved) product?
Key Independent Variables. Our research uses three
key independent variables: (i) social capitalsocial interactionfrom the factor analysis of social capital items
at the regional level; (ii) R&D intensity, a firm-level variable reflecting internal R&D efforts, measured by the
percentage of sales revenue spent on R&D; (iii) external
R&D acquisition, captured by externally acquired R&D
as a percentage of firm sales.
Firm-Specific Control Variables. In line with the existing literature, we control for the interaction between
R&D intensity and externally acquired R&D (Cassiman
and Veugelers 2006, Schmidt 2010). Moreover, although
the empirical research indicates that the advantages of
size for innovativeness are ambiguous, size is a commonly used variable in firm-level studies of innovation
(Cohen 1995). We measure firm size by the number
of employees. We control for firm age and for percentage of employees with degrees, which measures the
firms human capital. In addition, because the innovation literature shows that attention to user needs is
important for innovation success (Pavitt 1984, Slater and
Narver 1994), we use a dummy variable to describe the
degree of attention to customer satisfaction. Principal
activities are measured by three dummies for the Pavitt
(1984) sectors: supplier-dominated, scale-intensive, and
science-based, with specialized suppliers as the benchmark. To achieve a more fine-grained picture of the
industry, we control for average industry-level R&D
intensity, captured by R&D as a percentage of sales.
Region-Specific Control Variables. Region-specific
characteristics that might influence innovation include
political participation (social capitalpolitical participation); the percentage of the workforce with a science
and technology degree, which measures regional human
capital; and (public and private) regional expenditure on
R&D as a percentage of regional GDP. Population is
captured by the number of residents in a given region
in millions of people. To measure regional infrastructures, we include the control variables passengers by air
(the number of passengers embarked and disembarked
by air per every 100 inhabitants), port infrastructures
(tonnages of inbound and outbound goods transported
by sea per 100 inhabitants), and road infrastructures
(tonnages of inbound and outbound goods transported

Organization Science 23(1), pp. 177193, 2012 INFORMS

by road per 100 inhabitants). These infrastructure variables can be seen as a reflection of the general openness of a region (Gambardella et al. 2009). As proxies
for the regions economic activities, we use the number of firms over population, and the Herfindahl index
of industry concentration by region. The latter we measure using industry sales data pertaining to 38 industries
in each region. Furthermore, we add total tax paid on
productive activities (IRAP) by region expressed in billions of euros. Finally, we introduce dummy variables
for the 103 provinces and industry dummy variables for
the 96 industries classified according to a mix of twoand three-digit industry codes (aggregated such that each
industry includes at least five firms). Table 3 presents
descriptive statistics and correlations among the variables, two of which are quite high. This points to the
need to investigate problems related to multicollinearity,
as we discuss below.
Regression Analysis
Means of Estimation. Because our dependent variable
(product/no product innovation) is a binary variable, our
estimation is based on a logit model. The potential presence of endogeneity means we need to take account
of managements anticipation of performance outcomes
because the chosen strategy may lead to biased coefficient estimates. An endogeneity problem emerges when
there is a third, unobserved variable (e.g., whether or
not the firm in question is competent) that would affect
investment in both R&D and innovation. A competent
firm will be more likely to invest in internal and external R&D but will also be more likely to innovate.
Indeed, the management in competent firms will have
much stronger incentives to invest in external or internal R&D because an innovation outcome will be more
likely. If this is the case, part of the observed correlation
between R&D investments and innovation will be spurious because it could be ascribed to management selfselection, not to the fact that internal and external R&D
drive innovation (see Wooldridge 2002, pp. 5051).
It is possible to eliminate the endogeneity bias
using a two-stage instrumental variables (IVs) regression
approach, provided that the chosen instruments from
the first-stage model are not correlated with the unobserved variable representing managers self-selection in
the second stage of the model (Hamilton and Nickerson
2003). As we have two dependent variables in the first
stage, with a lower and an upper bound (R&D intensity
and the percentage of acquired R&D over total sales),
the fractional logit regression may be applied (Papke
and Wooldridge 1996). The model ensures that the predicted values of the dependent variable are in the interval
0 1. We use the predicted values from this first stage
of the model in lieu of the observed values for R&D
intensity and external R&D acquisition in the secondstage equation.

Correlation Matrix

Mean

002
000
001

267
10715
002

000
007

005
003

001 001
004
007

009
025
003

10

007 021
008 012
000 039

012 023
011
034

000

12

13

14

032
039
029
048
081
031 034 009 023

037

11

15

16

17

18

19

014
004
1566 001

006
001
034 004
000
008
000 002 009 001 003 002

001 002 001


001 003 001

005
000

006
004

005
001

008
001

000 003
000
000

007 006
000 001

008
002

758
010
094
005
011
001
009 004 001
007 031
052 003
011
006
030 042
10870 007 028 004
005 003 005
002 004 007
014 005 026 037 017 004
048 038
3011
004
021
003 017
000
017 002
005
009 030
047
048
095
082 005 020
032 046
034
003
000
000
040 001
006 004 001 004
001 004 007 005 002
001
003
000
008 006

001 001 009 013 043 024 008 002 023

015 003
005
013
001
005
000 001 001

009
012

003 006 043 004 014 004 004

000
001
012

001

006
004
004 001
005 002

008
006

001 022 001


004 007 001
000 007 001

005
002
008 004

045
002

005
002

026
040
013

016
008
035

177
036

016
005
000
024

006
033
003

023
005
078
042
010
002
003
009
016 005

010

045
008 002
086
018
006
084 002 010

282
042
32492
1872
732

050
076

S.D.

Note. Correlation coefficients above 004 are statistically significant at the 5% level.
a
Indicates a predicted variable from the first-step fractional response estimation.

1 Product innovation
0429
2 Social capital Social
0003
interaction
3 R&D intensity
0828
4 External R&D acquisition a
0781
5 Size
93619
6 Age
27467
7 Percentage of employees
5230
with a degree
8 Customer satisfaction
0713
9 Industry-level R&D intensity
0804
10 Social capital Political
0040
participation
11 Regional human capital
6274
12 Regional expenditure
0980
on innovation
13 Population
4900
14 Passengers by air
131663
15 No. of firms over
0087
the population
16 Herfindahl index of
0152
industry concentration
17 Road infrastructures
24071
18 Port infrastructures
72629
19 Tax paid
46327
20 Member of a
0133
commercial consortium
21 Labor flexibility
0085
22 Gearing ratio
2143

Table 3

002
002

20

000

21

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

Organization Science 23(1), pp. 177193, 2012 INFORMS

185

186

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

Because the instruments must not encompass the same


problem as the original regressor (Wooldridge 2002),
finding the most suitable instruments is crucial. The
instruments must be strong (i.e., must have an effect
in the first stage of the model) and valid. As a rule of
thumb, we would say that the instruments should not
be correlated with the dependent variable in the second stage of the model, although the true test of a valid
instrument is that it should not be correlated with the
part of the error term that is related to the problem
of endogeneity. We base our selection of instruments
on theoretical arguments and the Amemiya-Lee-Newey
minimum chi-square test, which tests whether the group
of instruments is valid according to the above. This tests
for whether it is necessary to accept (or reject) the entire
group of excluded instruments.
To measure R&D intensity, we identify three instruments likely to influence a firms investment in R&D but
not likely to influence innovation. The first is current
ratio, which is the current assets over current liabilities
and can be considered an indication of the firms ability
to meet its obligations. The higher the ratio, the more
liquid the company; the greater its financial strength,
the more likely the company will invest in R&D. However, it is difficult to predict the effect of a high current ratio on innovation because unpredictable events can
intervene in the relationship between liquidity and innovation outcome. Second is equity less noncurrent assets,
which is a measure of the difference between owners
equity and tangible fixed assets. A positive value for
this variable means that the firms investment in fixed
assets was financed using its own equity. The higher the
value of this variable, the more the firm draws on equity
for current activities. Consequently, it is reasonable to
believe high values for this variable mean that firms are
more likely to invest in R&D, but it may be difficult
to establish a direct effect on innovation. Third is the
gearing ratio, measured as owners equity over borrowed
funds, which provides a measure of the degree to which
the companys activities are financed by the owner or
creditor(s). When the gearing ratio is high, firms are
more vulnerable to business cycle downturns because
they must continue to pay off debts regardless of poor
sales. Thus, firms that present a high gearing ratio may
find it more difficult to invest in R&D because of their
greater vulnerability and resource constraints; however,
the effects of this variable on innovation are unclear.
To estimate external R&D acquisition, we use three
instruments. The first measures whether the firm is a
member of a commercial consortium, the argument here
being membership in a commercial consortium is indicative of more experience in cooperating with other organizations, which means that the variables are likely to
have a positive effect on external R&D acquisition without any direct link to the introduction of innovation.
Second is labor flexibility, measured as the number of

Organization Science 23(1), pp. 177193, 2012 INFORMS

employees on short-term contracts over the total number of employees, which describes a human resource
practice that may contribute to the utilization of external
knowledge, increasing the possibility that the firm will
forge new linkages outside its boundaries. Third is gearing ratio, which, as described above, does not seem to
have a clear effect on the propensity to innovate. However, it is conjectured that it has an effect on external
R&D acquisition: a high gearing ratio implies a resource
constraint and reduces the possibility that the firm will
acquire external R&D.
All the instruments in the first-stage regression model
are correlated (see Tables A.1 and A.2 in the appendix)
to the variables for which they instrument (R&D intensity and external R&D acquisition), which shows that
we have reasonably strong instruments. Because we are
using a fractional response logit-logit specification, we
have no perfect test for instrument validity. However,
the probit model with endogenous regressors allows us
to test the validity of our instruments. To get a better understanding of instrument validity, we use an IV
probit estimation of our equations (the IV probit model
produces results consistent with the fractional response
logit-logit model). The joint null hypothesis is that the
group of instruments is valid, i.e., the instruments are
uncorrelated with the error term in the structural equation, and the excluded instruments are correctly excluded
from the estimated equation. Using the three instruments
described above for R&D intensity, the Amemiya-LeeNewey minimum chi-square test statistic is 1.29, with
a corresponding p-value of 0.53. In other words, we
cannot reject the null hypothesis. However, with seemingly valid and strong instruments, we need to confirm
the existence of an endogeneity problem. In this case,
a Wald test for exogeneity does not reject the idea that
R&D intensity is exogenous to defining the probability to introduce a product innovation. Thus, we treat
R&D intensity as exogenous. Using the three instruments for external R&D acquisition, the Amemiya-LeeNewey minimum chi-square test static has a corresponding p-value of 0.31. This means that we cannot reject
the null hypothesis. In this case, the Wald test for exogeneity rejects the idea that acquisition of external R&D
is exogenous to determining the probability of being a
product innovator at the 5% level. Hence, we treat the
acquisition of external R&D as endogenous when considering product innovation.
Results
The results of the logit estimations for the second stage
of the procedure are reported in Table 4. Models I
and II represent our empirical model estimated with
all controls, including fixed effects for province and
the two- to three- digit industry for product innovation.4
These estimations suffer from serious multicollinearity
problems. In models I and II, the variance inflation factor

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

187

Organization Science 23(1), pp. 177193, 2012 INFORMS

Table 4

Results of the Logit Regressions for Product Innovation Using Instrumental Variables
Model I

Social capital Social


interaction
R&D intensity
External R&D acquisition
Social interaction
R&D intensity
Social interaction
External R&D acquisition a
R&D intensity
External R&D acquisition a
Size
Age
Percentage of employees
with a degree
Customer satisfaction
Supplier-dominated
Scale-intensive
Science-based
Specialized suppliers
Industry-level R&D intensity
Social capital Political
participation
Regional human capital
Regional expenditure
on innovation
Population
Passengers by air
No. of firms over
the population
Herfindahl index of
industry concentration
Road infrastructures
Port infrastructures
Tax paid
Industry dummies
Provinces dummies
Constant
No. of observations
Log likelihood
Chi-square
Pseudo R2
a

Model II

Model III

Model IV

Coef.

S.E.

Coef.

S.E.

Coef.

S.E.

16509

1585

18460

1537

0149

0076

0286

0079

0275

0080

5318
4676

0563
2033

6012
4739
2726

0580
2111
0569

5459
3332

0546
1930

5885
3235
2761

0550
2029
0528

6076
3385
2638

0561
2001
0548

4316

1994

4410

2162

4081

2141

4425

1187

4480

1202

0002
0001
0019

0000
0003
0007

0002
0001
0023

0000
0003
0008

0002
0000
0022

0000
0003
0008

0138

0115

0132

0116

0002
0001
0020

0000
0002
0007

Coef.

0002
0001
0020

S.E.

Model V

0000
0003
0007

Coef.

S.E.

0097
0075
0421 0120
0455 0151
0561
0313
Benchmark
0149
0108
0091
0071

0102
0075
0392 0122
0424 0152
0415
0313
Benchmark
0145
0109
0137
0073

0100
0075
0390 0122
0418 0152
0417
0312
Benchmark
0138
0109
0135
0073

0126
2401

0125
0149
0424 2703

0126
0431

0106
0722

0183 0137
1890
1212

0179
1855

0026
0070

0039
0198

0023
0034

0039
0209

0023
0034

0039
0209

0000
0008

0000
0012

0000
0002
1368

0000
0001
3228

0000
0001
0063

0000
0001
3399

0000
0001
0066

0000
0001
3398

0621

1895

1386

1895

1306

1894

0024

0550

0098

0557

0093

0557

0000
0009

0000
0012

1568
0005
0000
Yes
Yes

0163 1776
0005 0004
0000
0000
Yes
Yes

2,406
1,389
510.62
0155

2,406
1,370
548.57
0167

0157
0005
0000

2,413
1,443
409.25
0124

2,413
1,428
439.20
0133

2,413
1,426
443.42
0135

Indicates a predicted variable from the first-step fractional response estimation. Standard errors are in brackets.
p < 010; p < 005; p < 001; p < 0001 (two-tailed tests for controls, one-tailed tests for hypothesized variables).

(VIF) is higher than 9,500, which, among other things,


leads to greatly inflated parameters for the social capital
variables (partly expected, given that we have continuous
geographical and industry variables as well as detailed
geography and industry dummies). To circumvent these
multicollinearity problems, in the determinants of innovation, we dropped the geographic dummies and used
the Pavitt sectors to control for industry-level heterogeneity rather than the full set of industry fixed effects.
We also dropped three additional province-level variables. This reduces the VIF to well below the typically

recommended threshold of 10 (Belsley et al. 1980) in


models IIIV while still maintaining a sufficient number
of control variables. Adding any of the dropped variables
produces VIFs well in excess of 10.
Models IV provide support for Hypothesis 1 (ceteris
paribus, firms operating in regions with high levels of
social capital in terms of social interaction are more
likely to introduce product innovation), to the extent
that the social interaction component of social capital
is significant for explaining the likelihood of introducing product innovation. Furthermore, we find support for

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

188

Organization Science 23(1), pp. 177193, 2012 INFORMS

Hypothesis 2 (ceteris paribus, social capital and internal R&D spending are complementary in affecting the
likelihood of introducing product innovation). The interaction effect between the social interaction component
of social capital and R&D intensity is positive and significant (see model V). Because of the nonlinear nature
of the logit model, however, the marginal effect of an
interaction effect is not simply the coefficient of their
interaction (Hoetker 2007, Norton et al. 2004). In addition, because there are two additive terms, each of which
could be positive or negative, the interaction effect may
have different signs for different values of the covariates.
To deal with this complication, we apply a procedure
developed by Ai and Norton (2003) that computes correct magnitudes and standard errors for the interaction
effect. The vertical axis in Figure 1 presents the magnitude of the interaction effect, and in Figure 2 the vertical
axis shows the significance of the effect for each of the
observations. The horizontal axes show the models predicted probabilitytaking account of the effect of all

the covariatesthat the given firm is a product innovator. Figure 1 illustrates that the strongest interaction
effect occurs at the lower end of medium predicted levels of probability of being innovative (approximately 0.2
to 0.6), whereas the effect is less clear-cut for very low
and very high levels of the predicted probability of being
an innovator. Figure 2 also shows that in the majority of
cases the interaction effect is positive and significant at
the two-sided 5% level (in 89.0% of cases). The effect
is negative when the probability of being an innovator is
very high (the interaction effect is significant and negative in 116 of the 2,413 cases). Thus, although our findings are generally in line with Hypothesis 2, there is a
minority of observations to which it does not apply (see
the next section for a discussion of this finding).
Regarding Hypothesis 3 (ceteris paribus, the effectiveness of externally acquired R&D on the likelihood of
Figure 3

The Size Effect of the Interaction Between Social


Interaction and External R&D Acquisition

Interaction effect (percentage points)

Figure 1

Interaction effect (percentage points)

Interaction effects after logit


The Size Effect of the Interaction Between Social
Interaction and R&D Intensity

Interaction effects after logit


1.0

0.5

2.0

Incorrect marginal effect


1.5

1.0

0.5

0
0

Correct interaction effect


Incorrect marginal effect
0.4

0.6

0.2

0.4

0.6

0.8

1.0

Predicted probability that y = 1

0.5
0.2

Correct interaction effect

0.8

1.0

Note. External R&D acquisition is a predicted variable from the


first-step fractional response estimation.

Predicted probability that y = 1


Figure 4
Figure 2

The Significance of the Interaction Between Social


Interaction and R&D Intensity

The Significance of the Interaction Between Social


Interaction and External R&D Acquisition

z-Statistics of interaction effects after logit


10

z-Statistics of interaction effects after logit

z-Statistic

10

z-Statistic

0
0

0.2

0.4

0.6

0.8

1.0

Predicted probability that y = 1


0.2

0.4

0.6

0.8

Predicted probability that y = 1

1.0

Note. External R&D acquisition is a predicted variable from the


first-step fractional response estimation.

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation


Organization Science 23(1), pp. 177193, 2012 INFORMS

introducing product innovation is higher for firms operating in regions associated with high levels of social
capital), we can detect a positive and significant interaction effect on product innovation of the social interaction
component of social capital and the level of the firms
R&D that is externally acquired (see model V). Figure 3
illustrates that the strongest interaction effect occurs at
the medium predicted levels of the probability of being
innovative (approximately 0.3 to 0.6), whereas the effect
is less obvious at low and high levels of predicted probability of being an innovator (the effect is negative in only
15 of the 2,413 cases). In other words, for firms with
low predicted probability of being innovators, acquisition of external R&D combined with location in a region
with a high level of social capital makes little difference. Similarly, firms with a high predicted probability of being innovators are likely to innovate regardless.
Figure 4 illustrates that the interaction effect is positive
and significant in the majority of cases (95.4% of observations). This result provides rather general evidence of
the positive role of social capital in firms acquisitions
of external R&D.
Concerning the results for our control variables, it
should be noted that the interaction between R&D intensity and externally acquired R&D is negative and statistically significant. This finding is in contrast to some
contributions (e.g., Cassiman and Veugelers 2006) and
in line with other contributions (e.g., Laursen and Salter
2006, Schmidt 2010). As a robustness check, we ran
the model with process rather than product innovation
as the dependent variable. We found a significant direct
effect of social interaction on innovation. In addition,
social interaction positively and significantly moderates
the effect of external R&D acquisition on innovation,
whereas regional social interaction does not appear to
moderate the effect of R&D intensity on innovation.

Discussion and Conclusion

This paper demonstrates that geographically bound


social capital affects the innovative ability of firms in
21 Italian regions. We theorized that regional social
interaction helps shape product innovation through
localized connectivity and trust effects, and we found
empirical support for the significance of regional social
capital, in the form of social interaction, as an important driver of firm-level product innovation. In other
words, we provide evidence that location matters: firms
located in regions characterized by a high level of structural social capital in terms of social interaction display a higher propensity to innovate. This is consistent
with the literature on the role of geographical context
for firm performancee.g., industrial clusters, industrial districts, and territorial innovation systems (Asheim
and Gertler 2005, Brusco 1982, Porter 1990, Romanelli
and Khessina 2005, Tallman et al. 2004, Ter Wal and
Boschma 2011).

189

We found also that such regional social capital is complementary to internal R&D in affecting innovation and
that regional social capital enhances the functioning of
externally acquired R&D in the sense that it increases
the probability that external R&D leads to product innovation. However, our research also reveals an important nonlinearity: we found that among firms with a
high probability of being innovators (or leading-edge
organizations), high levels of regional social interaction mean that internal R&D is less likely to result in
innovation. We suggest that there may be two reasons
for this. The first could be that although social capital
makes inbound spillovers flow more easily, it also facilitates knowledge leakageand this could pose a problem, especially for the most advanced firms (Alccer
and Chung 2007). The second could be that leadingedge firms are often embedded in the local region but
are unable to find sufficient inspiration for knowledge
recombination in the local regional context. For these
few more advanced firms, there seems to be some
evidence of overembeddedness in the local region
(Uzzi 1997).
In the case of both internal and external R&D, we
found that the positive interaction effects on the probability of innovating are stronger for firms with a medium
probability of innovating. These medium-innovators
are generally considered to constitute the backbone of
the Italian economy, and the fact that they benefit the
most from social interaction effects is further empirical confirmation that this contextual knowledge factor is a key advantage for a local economy. It favors
greater participation in the innovation process of otherwise more disadvantaged firms, thus generating further
positive spillovers in the local context.
Our work provides two main contributions. First, we
have described explicit social mechanisms that underlie the notion of spillovers and that link a focal firm
to its geographical environment in the context of innovation. At the theoretical level, we have suggested that
the connectivity and trust effects of localized social capital play a central role in this regard. This study is
a first systematic attempt to demonstrate the existence
of Marshallian knowledge spillovers linked to localized
social capital from both theoretical and empirical angles.
From an empirical viewpoint, there is practically no
econometric analytical evidence on this issue. An important feature of our analysis is that it is based on a detailed
data set built from different data sources, constructed
to enable the selection of variables most appropriate for
our purposes. Our story and the suggested micromechanisms fit nicely with one of the main results of the
spillover literature: to absorb spillovers from the local
environment, a degree of absorptive firm-specific absorptive capacity is needed. Second, this paper adds to the
literature on the markets for technology (Arora et al.
2001, Fosfuri and Giarratana 2010), which so far has not

190

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

examined the role of geographically constrained institutions in explaining the degree of well-functioning of
these markets. We find support for the role of geography
in this regard but also for the more general point that
institutions not linked to the market in a narrow sense
play a central role in the functioning of the markets for
technology.
The findings of this study have implications for managerial practice. Whereas in regions with high levels of
social capital, this capital positively affects the effectiveness of externally sourced R&D and favors firm-level
innovation, in regions with low levels of social capital,
it is necessary for firms to invest more in accumulating
their own firm-specific social capital. For instance, they
can promote meetings, partnerships, and communication
with other firms and organizationsboth inside and outside the local region. For leading-edge firms, withinregion social interactioneven in a social capital-rich
regionmay not provide enough potential for knowledge recombination per se, making it essential for such
firms to establish social connections with firms and individuals located outside the region to fuel their investments in R&D. Finally, the theory and evidence provided
here show that the absorptive capacity hypothesis holds
also in the case of localized social capital: even when
firms have the advantage of location in a social capitalrich geographical environment, social capital is not a
free lunch. To properly benefit from localized social
capital, serious investment in ones own knowledge
generation is required.
This study has some limitations. We focus on the positive net effects of social capital. However, social capital
can also have negative consequences if the underlying
networks become too tight-knit. Prominent social capital theorists, such as Coleman (1988), stress the importance of dense networks as a prerequisite for the creation of social capital. However, dense networks may
incur penalties, e.g., the exclusion of outsiders, excess
claims on group members, restrictions on individual
freedoms, and downward leveling norms (see Portes
1998, p. 15, for a detailed discussion of these effects).
Although our measure of social capital can be considered a mix of strong and weak ties, we acknowledge
that it is not possible to identify these two types empirically. Although research that could separate these ties
would be of great value, not least to enable further
examination of the downsides of social capital in the
form of overembeddedness (Uzzi 1997), such an analysis would be extremely difficult at the relatively high
level of aggregation of the region.
Another limitation of this paper is that we focus on
only one period. Although we have corrected for endogeneity problems, controlled for the large number of
firm- and region-specific factors, and have a large sample of firms, the results of this study are based on
cross-sectional data. Also, although we rely on two data

Organization Science 23(1), pp. 177193, 2012 INFORMS

sources, collected at the firm and regional levels, we


do not know whether the externally acquired R&D was
purchased in the home region. Accordingly, the positive moderating effect of the regional social interaction
variable on the relationship between externally acquired
R&D and innovation may be related to the fact that a
high level of regional structural social capital makes the
acquiring firm better able to learn to deal with the process of outsourcing R&D and, at the same time, more
likely to be better connected socially to a selling firm
located in the home region. Future research should collect data on the geographic origins of acquired R&D to
disentangle these two effects.
Greater emphasis on how geographically bound social
capital enables and constrains managerial behavior in
other parts of the organization would appear to offer fertile ground for future research. In this paper, we focus
on R&D processes, an important element in the innovation process, but regional social capital might influence
the effectiveness of other of the firms external relations,
including formal collaboration between firms and external parties. Another avenue for future research would be
to separate out the effects of social capital and (internal and external) R&D spending on innovation at the
level of individual industries to explore possible industry variation. Insights from such research would provide
guidance for managers making decisions about how to
work with external parties in their environment.
Acknowledgments
The authors gratefully acknowledge comments from Gautam
Ahuja, Alfonso Gambardella, Lars Bo Jeppesen, Bart Leten,
Rekha Rao, two reviewers of this journal, the audiences at the
DRUID Conference 2007, and the Academy of Management
Conference 2008 on earlier versions of this paper. The authors
also thank Cristiano Zazzara and the research department at
Capitalia (now Unicredit) for the provision of the firm-level
data set used in this paper. The usual caveats apply. K. Laursen
acknowledges the financial support from the Danish Council
for Independent Research | Social Sciences [Grant 09-068739].

Appendix
Table A.1

First-Step Fractional Response Regressions


Explaining R&D Intensity

Explanatory variable
Current ratio
Equity less noncurrent assets
Gearing ratio
No. of observations
Log likelihood

Coeff.

S.E.

0620
0000
0006

0300
0000
0002

2,472
773

Notes. Estimates for the instruments are excluded in the second


step (parameters for the included instruments are not reported).
Standard errors are in brackets.

p < 010; p < 005; p < 001; p < 0001 (two-tailed tests).

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation


Organization Science 23(1), pp. 177193, 2012 INFORMS

Table A.2

First-Step Fractional Response Regressions


Explaining External R&D Acquisition

Explanatory variable
Member of a commercial consortium
Labor flexibility
Gearing ratio
No. of observations
Log likelihood

Coeff.

S.E.

0552
0967
0004
2,520
6413

0148
0314
0002

Notes. Standard errors are in brackets. Estimates for the instruments are excluded in the second step (parameters for the
included instruments are not reported).

p < 010; p < 005; p < 001; p < 0001 (two-tailed tests).

Endnotes
1

Although political participation is an interesting variable and


is considered very significant by an important part of the
regional social capital literature (Putnam et al. 1993), we do
not theorize about it because, given the ambiguous effect of
this variable on innovation outcomes, Italy currently does not
represent an empirical context that is appropriate for an examination of this potentially interesting issue; i.e., according to
observers, political parties in Italy are increasingly personal
machines (Calise 2000, p. 5), no longer accountable to members and activists sensitive to appeals for collective action
(Della Porta 2004).
2
Empirically, we control for the general openness of the
region, which to some extent may reflect out-of-region boundary spanning (see Gambardella et al. 2009).
3
By externally acquired R&D, we mean R&D conducted by
other than the acquiring firm, excluding the acquisition of
technologies.
4
Stata automatically drops the variable for the number of firms
over the population for reasons of collinearity.

References

Ai, C., E. C. Norton. 2003. Interaction terms in logit and probit models. Econom. Lett. 80(1) 123129.

191

Beugelsdijk, S., T. van Schaik. 2005. Differences in social capital


between 54 Western European regions. Regional Stud. 39(8)
10531064.
Brusco, S. 1982. The Emilian model: Productive decentralisation and
social integration. Cambridge J. Econom. 6(2) 167184.
Burt, R. S. 1992. Structural Holes: The Social Structure of Competition. Harvard University Press, Cambridge, MA.
Calise, M. 2000. Il Partito Personale. Laterza, Rome-Bari, Italiy.
Capitalia. 2006. Indagine sulle imprese Italiane: Rapporto sul sistema
produttivo e sulla politica industriale. Report, Capitalia, Rome.
Cassiman, B., R. Veugelers. 2006. In search of complementarity
in innovation strategy: Internal R&D and external knowledge
acquisition. Management Sci. 52(1) 6882.
Chesbrough, H. 2003. Open Innovation. Harvard University Press,
Cambridge, MA.
Cohen, W. M. 1995. Empirical studies of innovative activity.
P. Stoneman, ed. Handbook of the Economics of Innovation and
Technological Change. Blackwell, Oxford, UK, 342365.
Cohen, W. M., D. A. Levinthal. 1990. Absorptive capacity: A new
perspective of learning and innovation. Admin. Sci. Quart. 35(1)
128152.
Coleman, J. S. 1988. Social capital in the creation of human capital.
Amer. J. Sociol. 94(Supplement) S95S120.
Dahl, M. S., O. Sorenson. 2009. The embedded entrepreneur. Eur.
Management Rev. 6(3) 172181.
Della Porta, D. 2004. Political parties and corruption: Ten hypotheses
on five vicious circles. Crimes, Low Soc. Change 42(1) 3560.
Dyer, J. H., H. Singh. 1998. The relational view: Cooperative strategy
and sources of interorganizational competitive advantage. Acad.
Management Rev. 23(4) 660679.
Fleming, L., O. Sorenson. 2001. Technology as a complex adaptive system: Evidence from patent data. Res. Policy 30(7)
10191039.
Fosfuri, A., M. S. Giarratana. 2010. Trading under the Buttonwood:
A foreword to the markets for technology and ideas. Indust.
Corporate Change 19(3) 767773.

Alccer, J., W. Chung. 2007. Location strategies and knowledge


spillovers. Management Sci. 53(5) 760776.

Gambardella, A., M. S. Giarratana. 2010. Organizational attributes


and the distribution of rewards in a region: Managerial firms vs.
knowledge clusters. Organ. Sci. 21(2) 573586.

Almeida, P., B. Kogut. 1999. Localization of knowledge and the


mobility of engineers in regional networks. Management Sci.
45(7) 905917.

Gambardella, A., M. Mariani, S. Torrisi. 2009. How provincial is


your region? Openness and regional performance in Europe.
Regional Stud. 43(7) 935947.

Arora, A., A. Gambardella. 1990. Complementarities and external


linkages: The strategies of the large firms in biotechnology.
J. Indust. Econom. 38(4) 361379.

Granovetter, M. 1973. The strength of weak ties. Amer. J. Sociol.


78(6) 13601380.

Arora, A., A. Fosfuri, A. Gambardella. 2001. Markets for Technology: The Economics of Innovation and Corporate Strategy. MIT
Press, Cambridge, MA.
Asheim, B., M. S. Gertler. 2005. The geography of innovation:
Regional innovation systems. J. Fagerberg, D. C. Mowery, R. R.
Nelson, eds. The Oxford Handbook of Innovation. Oxford University Press, Oxford, UK, 291317.
Banfield, E. C. 1958. The Moral Basis of a Backward Society. Free
Press, New York.

Granovetter, M. 1985. Economic action and social structure: The


problem of embeddedness. Amer. J. Sociol. 91(3) 481510.
Guiso, L., P. Sapienza, L. Zingales. 2004. The role of social capital
in financial development. Amer. Econom. Rev. 94(3) 526556.
Gulati, R. 1995. Does familiarity breed trust? The implications of
repeated ties for contractual choice in alliances. Acad. Management J. 38(1) 85112.
Hamilton, B. H., J. A. Nickerson. 2003. Correcting for endogeneity
in strategic management research. Strategic Organ. 1(1) 5178.

Bell, G. G., A. Zaheer. 2007. Geography, networks, and knowledge


flow. Organ. Sci. 18(6) 955972.

Hill, C. W. L. 1995. National institutional structures, transaction cost


economizing and competitive advantage: The case of Japan.
Organ. Sci. 6(1) 119131.

Belsley, D. A., E. Kuh, R. E. Welsch. 1980. Regression Diagnostics:


Identifying Influential Data and Sources of Collinearity. John
Wiley & Sons, New York.

Hoetker, G. 2007. The use of logit and probit models in strategic


management research: Critical issues. Strategic Management J.
28(4) 331343.

192

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation

Jacobs, J. 1961. The Death and Life of Great American Cities. Random House, New York.
Jacobs, J. 1969. The Economy of Cities. Random House, New York.
Jaffe, A. B., M. Trajtenberg, R. Henderson. 1993. Geographic localization of knowledge spillovers as evidenced by patent citations.
Quart. J. Econom. 108(3) 577598.
Kalnins, A., W. Chung. 2006. Social capital, geography, and survival:
Gujarati immigrant entrepreneurs in the U.S. lodging industry.
Management Sci. 52(2) 233247.

Organization Science 23(1), pp. 177193, 2012 INFORMS

Papke, L. E., J. M. Wooldridge. 1996. Econometric methods for fractional response variables with an application to 401(k) plan participation rates. J. Appl. Econometrics 11(6) 619632.
Pavitt, K. L. R. 1984. Sectoral patterns of technical change: Towards
a taxonomy and a theory. Res. Policy 13(6) 343373.
Piore, M. J., C. F. Sabel. 1984. The Second Industrial Divide: Possibilities for Prosperity. Basic Books, New York.
Pisano, G. P. 1990. The R&D boundaries of the firm: An empirical
analysis. Admin. Sci. Quart. 35(1) 153176.

Knack, S., P. Keefer. 1997. Does social capital have an economic


payoff? A cross-country investigation. Quart. J. Econom. 112(4)
12511288.

Podsakoff, P. M., D. W. Organ. 1986. Self-reports in organizational research: Problems and prospects. J. Management 12(4)
531544.

Kogut, B., U. Zander. 1996. What firms do? Coordination, identity,


and learning. Organ. Sci. 7(5) 502518.

Porter, M. E. 1990. The Comparative Advantage of Nations. Free


Press, New York.

La Porta, R., F. Lopez-de-Silanes, A. Shleifer, R. W. Vishny.


1997. Legal determinants of external finance. J. Finance 52(3)
11311150.

Portes, A. 1998. Social capital: Its origins and applications in modern


sociology. Annual Rev. Sociol. 24 124.

Landry, R., N. Amara, M. Lamari. 2002. Does social capital determine


innovation? To what extent? Tech. Forecasting Soc. Change
69(7) 681701.

Portes, A., P. Landolt. 1996. The downside of social capital. Amer.


Prospect 26(MayJune) 1821.
Putnam, R. D. 2000. Bowling Alone: The Collapse and Revival of
American Community. Simon & Schuster, New York.

Lane, P. J., M. Lubatkin. 1998. Relative absorptive capacity and


interorganizational learning. Strategic Management J. 19(5)
461477.

Putnam, R. D., R. Leonardi, R. Y. Nanetti. 1993. Making Democracy


Work: Civic Traditions in Modern Italy. Princeton University
Press, Princeton, NJ.

Larson, A. 1992. Network dyads in entrepreneurial settings: A study


of the governance of exchange relationships. Admin. Sci. Quart.
37(1) 76104.

Ring, P. S., A. H. van de Ven. 1994. Developmental processes of


cooperative interorganizational relationships. Acad. Management
Rev. 19(1) 90118.

Laursen, K., A. J. Salter. 2006. Open for innovation: The role of


openness in explaining innovative performance among UK manufacturing firms. Strategic Management J. 27(2) 131150.

Robins, J., M. F. Wiersema. 1995. A resource-based approach to the


multibusiness firm: Empirical analysis of portfolio interrelationships and corporate financial performance. Strategic Management J. 16(4) 277299.

Lindenberg, S. 1996. Constitutionalism versus relationalism: Two


views of rational choice sociology. J. Clark, ed. James S. Coleman. Falmer Press, London, 299312.
Lundvall, B.-., ed. 1992. National Systems of Innovation: Towards a
Theory of Innovation and Interactive Learning. Pinter Publishers, London.
Marshall, A. 1890. Principles of Economics, 1st ed. Macmillan and
Co., London.
McGahan, A. M, M. E. Porter. 1997. How much does industry matter,
really? Strategic Management J. 18(Special issue) 1530.

Romanelli, E., O. M. Khessina. 2005. Regional industrial identity:


Cluster configurations and economic development. Organ. Sci.
16(4) 344358.
Rosenkopf, L., P. Almeida. 2003. Overcoming local search through
alliances and mobility. Management Sci. 49(6) 751766.
Rosenkopf, L., A. Nerkar. 2001. Beyond local research: Boundaryspanning, exploration, and impact in the optical disk industry.
Strategic Management J. 22(4) 287306.
Rosenthal, S. S., W. C. Strange. 2003. Geography, industrial organization, and agglomeration. Rev. Econom. Statist. 85(2) 377393.

Morgan, K. 2004. The exaggerated death of geography: Learning,


proximity and territorial innovation systems. J. Econom. Geography 4(1) 321.

Rothaermel, F. T., M. T. Alexandre. 2009. Ambidexterity in technology sourcing: The moderating role of absorptive capacity.
Organ. Sci. 20(4) 759780.

Nahapiet, J., S. Ghoshal. 1998. Social capital, intellectual capital,


and the organizational advantage. Acad. Management Rev. 23(2)
242266.

Rothwell, R., C. Freeman, A. Horlsey, V. T. P. Jervis, A. B. Robertson, J. Townsend. 1974. SAPPHO updatedProject SAPPHO
phase II. Res. Policy 3(3) 258291.

Nelson, R. R., S. G. Winter. 1982. An Evolutionary Theory of Economic Change. Harvard University Press, Cambridge, MA.

Salancik, G. R., J. Pfeffer. 1977. An examination of need-satisfaction


models of job attitudes. Admin. Sci. Quart. 22(3) 427456.

North, D. C. 1989. Institutions and economic growth: An historical


introduction. World Dev. 17(9) 13191332.

Saxenian, A. 1994. Regional Advantage: Culture and Competition


in Silicon Valley and Route 128. Harvard University Press,
Cambridge, MA.

Norton, E. C., H. Wang, C. Ai. 2004. Computing interaction effects


and standard errors in logit and probit models. Stata J. 4(2)
154167.

Schmidt, T. 2010. Absorptive capacityOne size fits all? Managerial


Decision Econom. 31(1) 118.

Owen-Smith, J., W. W. Powell. 2004. Knowledge networks as channels and conduits: The effects of spillovers in the Boston
biotechnology community. Organ. Sci. 15(1) 521.

Shan, W., G. Walker, B. Kogut. 1994. Interfirm cooperation and


startup innovation in the biotechnology industry. Strategic Management J. 15(5) 387394.

Oxley, J. E. 1997. Appropriability hazards and governance in strategic


alliances: A transaction cost approach. J. Law, Econom., Organ.
13(2) 387409.

Slater, S. F., J. C. Narver. 1994. Does competitive environment moderate market orientation-performance relationship? J. Marketing
58(1) 4655.

Laursen, Masciarelli, and Prencipe: Localized Social Capital Affects Innovation


Organization Science 23(1), pp. 177193, 2012 INFORMS

193

Sorenson, O. 2003. Social networks and industrial geography. J. Evol.


Econom. 13(5) 513527.

Williamson, O. 1985. The Economic Institutions of Capitalism. Free


Press, New York.

Sorenson, O., P. G. Audia. 2000. The social structure of


entrepreneurial Activity: Geographic concentration of footwear
production in the United States, 19401989. Amer. J. Sociol.
106(2) 424462.

Woolcock, M., D. Narayan. 2000. Social capital: Implications for


development theory, research, and policy. World Bank Res.
Observer 15(2) 225249.

Storper, M., A. J. Venables. 2004. Buzz: Face-to-face contact and the


urban economy. J. Econom. Geography 4(4) 351370.
Stuart, T., O. Sorenson. 2003. The geography of opportunity: Spatial
heterogeneity in founding rates and the performance of biotechnology firms. Res. Policy 32(2) 229253.
Szulanski, G. 1996. Exploiting internal stickiness: Impediments to the
transfer of best practice. Strategic Management J. 17(Special
issue) 2743.
Tallman, S., M. Jenkins, N. Henry, S. Pinch. 2004. Knowledge, clusters and competitive advantage. Acad. Management Rev. 29(2)
258271.
Teece, D. J. 1986. Profiting from technological innovation: Implications for integration collaboration, licensing and public policy.
Res. Policy 15(6) 285305.
Teece, D. J. 1988. Technological change and the nature of the firm.
G. Dosi, ed. Technological Change and Economic Theory. Pinter
Publishers, London, 256281.
Ter Wal, A. L. J., R. A. Boschma. 2011. Co-evolution of firms, industries and networks in space. Regional Stud. 45(7) 919933.
Tsai, W., S. Ghoshal. 1998. Social capital and value creation: The role
of intrafirm networks. Acad. Management J. 41(4) 464476.
Uzzi, B. 1997. Social structure and competition in interfirm networks:
The paradox of embeddedness. Admin. Sci. Quart. 42(1) 3567.
von Hippel, E. 1988. The Sources of Innovation. Oxford University
Press, New York.
von Hippel, E. 1994. Sticky information and the locus of problem solving: Implications for innovation. Management Sci. 40(4)
429439.
von Hippel, E. 2005. Democratizing Innovation. MIT Press,
Cambridge, MA.
Williamson, O. 1979. Transaction-cost economics: The governance of
contractual relations. J. Law Econom. 22(2) 233261.

Wooldridge, J. M. 2002. Econometric Analysis of Cross Section and


Panel Data. MIT Press, Cambridge, MA.
Yli-Renko, H., A. Autio, H. J. Sapienza. 2001. Social capital, knowledge acquisition, and knowledge exploitation in
young technology-based firms. Strategic Management J. 22(67)
587613.
Zahra, S. A., R. D. Ireland, M. A. Hitt. 2000. International expansion
by new venture firms: International diversity, mode of market
entry, technological learning, and performance. Acad. Management J. 43(5) 925950.
Zak, P. J., S. Knack. 2001. Trust and growth. Econom. J. 111(470)
295321.

Keld Laursen is a professor of the Economics and Management of Innovation at Copenhagen Business School. He earned
his doctorate in international economics from Aalborg University. His current research focuses on open and distributed
innovation processes and on organizational practices and innovation outcomes. He serves on several editorial boards.
Francesca Masciarelli is a research fellow at the University
G. dAnnunzio. She received her doctorate from the University
of Trento. Her research interests include social capital, strategy
and management of innovation and international business, with
particular emphasis on the implications of social capital on
firms competitiveness.
Andrea Prencipe is a professor of innovation (business and
management) at the University G. dAnnunzio. He received
his doctorate from SPRU (University of Sussex), where he
is an honorary professor. His research interests include the
implications of modular design strategies on the division and
coordination of labor; organizational learning in project-based
organizations; the relationships between social capital, innovation and internationalization processes.

CORRECTION
In this article, Regions Matter: How Localized Social Capital Affects Innovation and External Knowledge
Acquisition by Keld Laursen, Francesca Masciarelli, and Andrea Prencipe (first published in Articles in
Advance, May 17, 2011, Organization Science, DOI: 10.1287/orsc.1110.0650), Stuart and Sorenson (2003,
p. 232) were misquoted. The quote has been corrected to read as follows: When people with common professional interests cluster in physical space, informal social and professional networks emerge and serve to
disseminate information.

Оценить